A New Adaptive Method to Identify Inrush Using Grille Fractal

This paper presents a novel scheme based on grille fractal for distinguishing between inrush currents and internal fault currents in transformers, which is derived from the singularity detection in the waveforms. First, differential currents are analyzed on the basis of grille fractal (GF). Noises are then filtered by use of the adaptive generalized morphological filter (AGMF). In a total of 158 experimental cases, the inrush current is characterized by the drastic variation, but the faulty current is kept almost constant. The experimental results indicate that the proposed technique is stable and reliable during discrimination between internal faults and magnetizing inrush currents, and saturation of current transformers (CTs).